(3) In others there are levers or processes by which the anthers are mechanically brought down on to the head or back of an insect entering the flower, in such a position as to be carried to the stigma of the next flower it visits. This may be well seen in many species of Salvia and Erica.
(4) In some there is a sticky secretion which, getting on to the proboscis of an insect, carries away the pollen, and applies it to the stigma of another flower. This occurs in our common milkwort (Polygala vulgaris).
(5) In papilionaceous plants there are many complex adjustments, such as the squeezing out of pollen from a receptacle on to an insect, as in Lotus corniculatus, or the sudden springing out and exploding of the anthers so as thoroughly to dust the insect, as in Medicago falcata, this occurring after the stigma has touched the insect and taken off some pollen from the last flower.
(6) Some flowers or spathes form closed boxes in which insects find themselves entrapped, and when they have fertilised the flower, the fringe of hairs opens and allows them to escape. This occurs in many species of Arum and Aristolochia.
(7) Still more remarkable are the traps in the flower of Asclepias which catch flies, butterflies, and wasps by the legs, and the wonderfully complex arrangements of the orchids. One of these, our common Orchis pyramidalis, may be briefly described to show how varied and beautiful are the arrangements to secure cross-fertilisation. The broad trifid lip of the flower offers a support to the moth which is attracted by its sweet odour, and two ridges at the base guide the proboscis with certainty to the narrow entrance of the nectary. When the proboscis has reached the end of the spur, its basal portion depresses the little hinged rostellum that covers the saddle-shaped sticky glands to which the pollen masses (pollinia) are attached. On the proboscis being withdrawn, the two pollinia stand erect and parallel, firmly attached to the proboscis. In this position, however, they would be useless, as they would miss the stigmatic surface of the next flower visited by the moth. But as soon as the proboscis is withdrawn, the two pollen masses begin to diverge till they are exactly as far apart as are the stigmas of the flower; and then commences a second move-
